Skip to content
IRIS NEURALContact

Solutions by typeInfrastructure integrity

How do you know which row has stopped producing?

A whole field and nobody walking it

One in the afternoon at the coastal solar farm. Every row is turned the same way, following the sun, except one: the middle row has stayed facing the other way and is showing the back of its modules. It’s been like that since nine fifteen this morning. Nobody on site knows, because there’s nobody on site: it’s eighty acres and the next visit is scheduled for next Thursday.

9 cameras across 7 sectors

This is how you ask for it

«Tell me if any tracker row ends up at a different angle from its neighbors for more than an hour.»

What a solar farm loses, it loses quietly. At the end of the month there is a total, and a total doesn’t tell you which row. Between the day something breaks and the day someone links it to a specific row, weeks of generation can go by that never come back. And with hail it is worse: to make a claim you need the date and the before-and-after image, and if you didn’t have that, it can’t be manufactured afterward.

You decide once

You decide this once. You say which rows are watched, how long one may sit out of line before it counts, what hours it applies — here, daylight only — and who gets told: if a row holds a different angle for more than an hour, alert the maintenance team with the image of the field and the time it fell out of step. IRIS checks that every day, including the days with no visit scheduled, which on a solar farm is almost every day.

A solar farm is the opposite of a factory. In a factory people walk past the machines all day; on a farm there are hundreds of acres of hardware running on their own, and what tells you something has broken is usually a number at the end of the month. A number that is also a total: it says output was lower, it doesn’t say which row. And whoever goes looking for the row has to walk the field, which is precisely the job nobody has time to do every week.

What needs seeing, though, has a very recognisable shape from a camera. A tracker row facing the opposite way from all the others at midday. Three rows that look whitish instead of deep blue because the glass shattered in last night’s hail. All of it is visible from the same place a person walking the track would see it, with the difference that the camera is looking on Tuesday afternoon and the person isn’t.

IRIS looks at the field row by row and says when one has stopped resembling its neighbors, or itself last week. It alerts with the image, with the time it stopped following the sun and with the comparison alongside. That pair of dated images is also what you need later to claim for damage, and it’s exactly what you never have when the damage turns up three weeks late.

What changes

Before, the field was walked every two or three weeks, and whatever failed between visits was found on the next one, or on the bill. Now, the day a row ends up facing the wrong way, the alert arrives with the time and the image, and whoever drives out already knows which row and what to bring. The round isn’t abolished, it’s sharpened: instead of walking eighty acres looking, you go and see what you already know is wrong.

Where we have seen it

  • The crossroads every transfer passenger goes through. On the left, a wall of flight boards with people standing in front of it looking for their gate, and a row of seats in use. In the middle, the signs that split the flow: one toward transfers and one toward the two gate branches. At the back, the escalators going up and a busy corridor. Across the floor, a continuous fan of people with cabin bags. IRIS counts how many people take each branch and measures how long they stand still in front of the board before deciding which way to go.

    CAM-586Three minutes standing at the board isn’t walking slowly

    «Tell me if the average time stood at the board goes over five minutes for a whole hour.»

    See this cameraAirports

  • Afternoon shift on the trim line. A silver body moves along with the tailgate and the hood open. A branch of the wiring harness hangs loose down the left flank and a white connector has been left out of its housing: it isn’t mated. The operator working right in front of it will carry on fitting, and the trim panel will cover that spot at the next station. A blue cart of spare harnesses waits alongside. IRIS sees the loose connector while it can still be seen.

    CAM-225A connector left unmated on the line

    «Tell me if a connector is left unmated on the flank before the trim panel goes on.»

    See this cameraIndustry

  • Mid-morning in the tool room. Wrenches, hammers and pliers hang on the blue board, each one over its marked place. One outline is empty: the red marking is there, the tool isn’t. Beside it, the changeover cart carries a red foam tray with unfilled pockets. Two operators are working at the bench. Nobody has noticed: the gap only shows if someone looks at the whole board.

    CAM-250One outline on the board is empty: a tool is missing

    «Tell me when a board outline is left without its tool at the end of the shift.»

    See this cameraIndustry

  • Aseptic filling area. A panel of the isolator is open and an operator has put an arm through the gap: the gloved hand is touching the vials directly, inside the barrier. On either side, the isolator glove ports are still in place, unused. Behind, the filling carousel turns with the vials moving on. Further back, another person is working at the control panel with their back turned, unaware of what’s happening.

    CAM-294An arm reaching into the isolator through the open panel

    «Alert me the moment anyone opens an isolator panel and reaches in where the vials are.»

    See this cameraIndustry

  • A workshop in mid-morning. On the wall, a board with every tool hanging in its own slot: wrenches in size order, screwdrivers, pliers, hammers, files, a saw and a level. Below it runs a bank of labeled drawers. An officer, seen from behind, is checking the board with the sheet in his hand, and three of the people held there are working at the wooden benches. In this picture every tool is present: none is missing, so there’s nothing to attend to and the point shows in gold.

    CAM-115The tool board in the workshop

    «Tell me if any tool is missing from the board when the workshop closes, and tell me which slot and since what time it has been empty.»

    See this cameraThe prison

  • A variable message sign above the roadway is showing the word “DANGER” with some of its LEDs burned out: the letters in the middle have pieces missing and can’t be read from a car.

    CAM-70Variable message sign

    «Tell me if a sign looks broken, dark or with the message incomplete when seen from the road.»

    See this cameraThe highway

  • Midday at the coastal solar plant. Every row is turned the same way, following the sun, except one: the row in the middle has stayed facing the opposite way and shows the backs of the modules and the bare frame. Its neighbors, on both sides, are aligned with each other. The whole comparison fits inside one frame, which is why the difference is obvious at a glance.

    CAM-359One tracker row is facing the opposite way to all the others

    «Tell me if any tracker row stays at a different orientation from its neighbors for more than an hour.»

    See this cameraIndustry

  • A solar field on the hillside, mid-morning. Eight rows of panels facing the sea. The upper ones are perfect, deep blue and even. Down at the bottom, three rows look whitish: the glass is chipped and cracked, there are modules split and shifted out of place and, in the middle, a gap where a panel is missing and the bare frame shows. From this same camera, yesterday they all looked the same.

    CAM-354Three rows of panels with the glass shattered

    «Tell me if broken, dirty or displaced panels show up in any row of the field.»

    See this cameraIndustry

What it doesn’t do

IRIS does not measure generation and does not replace the inverters or the energy metering system. It does not tell you how many kilowatts were lost: it tells you that row is not like the others, and since when.

And it cannot see what is not visible. A cell cracked inside, a connection that has come loose, a degradation that does not change the color: all of that is out of reach. Thermography and electrical testing are still needed for those.

Questions

Doesn’t the inverter already catch this?
The inverter does detect that part of the field is doing worse, and that’s already a lot. What it doesn’t give you is the cause or the exact place: it can’t tell a row turned the wrong way from a broken module or a new shadow. The camera shows what is happening to that row, and with that the crew sets out knowing what to bring. The two work well together: one says something’s wrong, the other says what it looks like.
Do you need a lot of cameras for a large field?
You don’t need one per row, but one for everything won’t do either. From a high point, a single camera covers a good number of rows and sees differences in angle and color perfectly well. Whatever falls outside the frame isn’t covered, and that gets said up front rather than discovered the day of the hailstorm. It’s done the other way around from usual: first you decide which parts of the field matter, then you decide how many cameras it takes to see them.
Is the image any use for an insurance claim?
Yes, and it’s usually the piece that was missing. With the before image and the after image, both dated and from the same camera, you can show when it was fine and when it stopped being fine. That’s exactly what a loss adjuster asks for and exactly what you don’t have when the damage is found weeks later. The adjuster and the insurer still decide the settlement; IRIS only supplies the evidence of when it happened.
And at night, when the field isn’t producing?
At night half the rules switch themselves off, because a row can’t follow the sun when there’s no sun. What doesn’t turn off is the other half: nobody should be on the site. A solar farm at night is a large empty enclosure with a lot of copper in it, and the rule for those hours is a different one from the midday rule. You write both, and each applies in its own window, with nobody having to change anything at dusk.

See all the questions

Demo · IRIS at work

A new case each time. Always the same screen.

It plays by itself and shows one case from start to finish. It’s a summary: the real screen has far more functions than fit in this box.

  1. Cameras on the map
  2. You draw an area
  3. What passed shows up
IRIS NEURAL
30ES
IRISDirectoGrabacionesGISIncidencias3996SituaciónCasosLPRAutomatizacionesIRIS DATA
Ask a question or make a request…Send
IRIS · Infinity Neural

Not seeing yours here?

These words are the ones we have built, not the ones IRIS understands. The list falls short on purpose: you ask for it in a sentence, and sentences don’t run out. Tell us what is in front of your cameras and we’ll tell you whether it understands it, or not yet.

See all solutionsTell us what you see

And the best part

Video tells you what happened.An alert reaches you while it’s happening.

Reviewing the recording helps you explain. Hearing about it at the time helps you get there. Two different things, and until now you only had the first.

Tell us your problem and we’ll say whether IRIS understands it, or not yet.

A person answers, the same working day.